How to get cited by AI comes down to how each paragraph is shaped. Here is the writing framework we use for every Orkkid client, with the patterns AI engines quote.
Most content never gets cited by AI engines, and the reason is rarely quality or length.
If you want to know how to get cited by AI, start with the shape of your paragraphs. AI engines lift self-contained passages that open with a direct answer and carry specific, checkable facts, so the same expertise written in that shape gets quoted far more often.
ChatGPT, Perplexity, Gemini, and Google AI Overviews all do something specific when they answer a question. They scan candidate sources. They lift a self-contained passage. They attribute it. The unit of citation is the paragraph, so that is where the work happens.
Write paragraphs the AI can lift cleanly, and you get cited. Write paragraphs that need three sentences of context to make sense, and you do not.
This is the writing framework we use for every Orkkid client. It is the content side of generative engine optimization (GEO). For the wider strategy, pair it with our guide to what generative engine optimization is.
TL;DR
- In our October 2026 test of 120 AI answers recommending local businesses across eight cities, 60.8% of answers quoted a specific review count or star rating, and industry review sites supplied 39.6% of the sources the AIs cited, against 16% from the businesses' own websites.
- To write content AI will cite, shape paragraphs the engines can lift cleanly. The unit of citation is the paragraph, not the page.
- Every citable passage has four properties: self-contained meaning, a definitive opening sentence, specific verifiable facts, and schema-backed structure.
- The paragraph shapes AI cites most are the definition pattern, the list-with-rationale pattern, the question-as-heading pattern, and the comparison pattern.
- The highest-leverage edits to existing content are removals: cut transitional sentences, pronouns that point to earlier paragraphs, vague qualifiers, rhetorical questions, and "we" without a named subject.
- Add your brand name, specific numbers, dates, named tools, and geographic detail in place of generic claims.
- Paragraphs of 60 to 120 words are the most citable. Test them with the lift test, the context-free test, and the first-sentence test.
- Traditional SEO rewards articles. GEO rewards paragraphs. A #1-ranked article can still produce zero AI citations if its paragraphs are not self-contained.
The four properties of a citable passage
Every paragraph that gets cited shares four properties. Miss any one and the AI keeps scrolling.
1. Self-contained meaning
The paragraph makes complete sense without reading the paragraphs around it.
If your reader needs the previous section to understand a sentence, the AI also needs that section to make sense of it. AI engines do not pull two-paragraph quotes. They pull one. So when they hit a paragraph that needs context, they keep looking for something cleaner.
The fix is to repeat the noun once per paragraph. If you wrote "they offer this in three packages" and "they" refers to a business named two paragraphs earlier, an AI lifting that paragraph cannot use it. Rewrite it as "Acme Moving offers this in three packages" and it becomes liftable.
2. Definitive opening sentence
The first sentence of every paragraph should make a definitive claim that sums up the paragraph.
Bad opener: "There are several things to consider when choosing a moving company."
Good opener: "The three things that matter most when choosing a moving company are licensing, insurance coverage, and a written estimate."
The good opener is itself a citable answer. The bad opener is a setup. AI engines pull answers, not setups.
When you write, ask yourself: if an AI cited only the first sentence of this paragraph, would the reader learn something useful? If not, rewrite it.
3. Specific facts, not generic claims
AI engines prefer paragraphs with specific, verifiable facts (numbers, dates, named entities, places) over paragraphs with generic descriptions.
Generic: "Our service is fast and reliable."
Specific: "Most moving estimates are sent within one business day, and we have completed over 800 moves since 2021."
The specific paragraph is more likely to be cited because the AI can confidently attribute the claim. Generic claims could come from anywhere. Specific claims belong to your brand.
4. Schema-backed structure
The paragraph is reinforced by schema markup on the page that tells the AI exactly what the page is about.
A page with Service schema for long-distance moving, plus a paragraph saying "Acme Moving handles long-distance moves to every state, with a two-day pickup window", reinforces both signals. The AI gets a structured fact (Service: long-distance moving, areaServed: United States) plus a citable paragraph that matches it. The combination is far more likely to be lifted than either signal alone.
You can build the core types with our free schema generator, and add FAQ markup with the FAQ schema generator.
| Property | What it means | Quick test |
|---|---|---|
| Self-contained meaning | Makes sense without the paragraphs around it | Does it name its subject instead of saying "they" or "this"? |
| Definitive opening sentence | First sentence is the answer, not a setup | Would the first sentence alone be useful? |
| Specific facts | Numbers, dates, names, places | Could a competitor copy the sentence word for word? If yes, it is too generic. |
| Schema-backed structure | Page markup matches the paragraph's claim | Does the schema describe the same service and area? |
The paragraph patterns AI loves
A few paragraph shapes get cited far more than others. Use these as templates.
Definition pattern
Open with the term, follow with a one-sentence definition, then a short clarifier.
Example:
Citation share is the percentage of relevant AI prompts that mention your business by name. It is calculated by running a fixed prompt panel monthly and counting the prompts that named you. It is the leading indicator of GEO performance.
AI engines lift this pattern often because it answers the question in two sentences.
List-with-rationale pattern
Three-to-five-item lists where each item has a short reason.
Example:
The three biggest mistakes wedding photographers make on their websites are: galleries with no alt text (search engines and AI cannot tell what the images show), one generic "areas we serve" page instead of region pages (couples search by region), and hiding price ranges (couples ask AI "how much does a wedding photographer cost" and the AI quotes whoever answered).
This gets cited for "common mistakes" or "what to avoid" questions because it is structured, scannable, and specific.
Question-as-heading pattern
Use customer questions as H2 or H3 headings, then answer in 60 to 100 words.
Example:
## How long does a GEO program take to show citation gains?
Most service businesses see their first measurable citation share movement within 4 to 8 weeks if they already have some Google authority. New domains take 3 to 6 months because AI engines need time to learn the brand exists and trust it. The trend matters more than any single reading. Consistent monthly tracking is what separates programs that work from programs that quit at week eight.
For how to measure this, see how to track ChatGPT citations.
Comparison pattern
X vs Y framing where X and Y are real options the reader is weighing.
Example:
A specialist GEO agency will deliver tighter citation tracking and faster monthly iteration. A full-service digital agency with GEO bolted on will deliver broader media reach but slower GEO-specific work. For most service businesses on a modest budget, the specialist route compounds faster.
This pattern is the backbone of our GEO vs SEO comparison and our guide on how to choose a GEO agency.
→ Free free GEO audit: see whether ChatGPT and Google AI recommend your business, and who they name instead. Check my AI visibility.
What to remove from your existing content
If you are auditing an existing site for citability, the highest-leverage edits are removals, not additions.
Remove transitional sentences. "As we mentioned earlier", "In the next section", "First, let us look at..." None of these survive an AI lift. Cut them.
Remove pronouns that point to earlier paragraphs. "It", "they", "this approach" become useless in a paragraph quoted out of context. Replace them with the noun.
Remove vague qualifiers. "Many businesses", "some studies show", "it is generally agreed". These mark a paragraph as opinion rather than fact, and AI engines prefer fact.
Remove rhetorical questions used as section openers. "Have you ever wondered why...?" An AI cannot lift a question and present it as an answer.
Remove "we" without a named subject. "We help businesses grow" is uncitable. "Orkkid is a GEO and AI SEO agency that gets businesses recommended by ChatGPT, Perplexity and Google AI" is.
What to add instead
The reverse of the cuts above:
- Add the brand name in every paragraph that describes your offer. Not in every sentence, but at least once.
- Add specific numbers wherever the copy said "many" or "several".
- Add dates to vague claims. "In business since 2008" is more citable than "we have been around for years".
- Add named tools and platforms. "We use DataForSEO" is more citable than "we use industry-leading tools".
- Add geographic detail. "Across the metro area and within 50 miles of downtown" is more citable than "locally".
For local service businesses, geographic detail opens up a whole class of questions that generic competitors never get named for. Our GEO services page shows how we build that into every site.
Find out if AI recommends your business
What AI actually cited and repeated in 120 local recommendations
The framework above shapes your own pages. To see how much weight those pages carry, we asked five AI assistants (ChatGPT, Perplexity, Gemini, Claude and Google AI Mode) to recommend moving companies, wedding photographers and solar installers in eight cities in October 2026, then classified every source they cited and searched every answer for the facts they repeated. The full results are in our AI recommendation study.
Your own website is one source among several
Where each AI gets its recommendations
Share of cited sources by type
Industry review sites are the biggest single source: 39.6% of everything the five AIs cited.
ChatGPT
40%19%10%10%9%Perplexity
47%21%18%Gemini
30%33%12%12%Claude
46%15%12%12%Google AI Mode
25%10%12%29%
Source: Orkkid AI Recommendation Study, October 2026 (120 answers, unique domains per answer)
Industry review sites such as moveBuddha, EnergySage, WeddingWire and The Knot made up 39.6% of everything the five AIs cited. Businesses' own websites came second at 16%, followed by best-of lists at 11.7% and media and publishers at 11.3%. The mix varies by engine: Gemini drew 32.7% of its sources from business websites, while Perplexity drew only 3.9% from them and 46.7% from industry review sites.
These were the 15 domains cited most often across all 120 answers:
| Domain | Answers citing it (of 120) | Share | Type of site | Industries where it was cited |
|---|---|---|---|---|
| reddit.com | 31 | 25.8% | Reddit and social | Moving, wedding, solar |
| consumeraffairs.com | 30 | 25% | Industry review site | Moving, solar |
| energysage.com | 26 | 21.7% | Industry review site | Solar |
| weddingwire.com | 21 | 17.5% | Industry review site | Wedding |
| movebuddha.com | 19 | 15.8% | Industry review site | Moving |
| theknot.com | 18 | 15% | Industry review site | Wedding |
| freightwaves.com | 17 | 14.2% | Media and publishers | Moving |
| yelp.com | 16 | 13.3% | Directory | Moving, wedding, solar |
| zola.com | 16 | 13.3% | Industry review site | Wedding |
| google.com | 14 | 11.7% | Moving, wedding, solar | |
| forbes.com | 12 | 10% | Media and publishers | Moving, solar |
| extraspace.com | 12 | 10% | Other (storage company blog) | Moving |
| serviceagent.ai | 11 | 9.2% | Best-of list | Solar |
| expertise.com | 9 | 7.5% | Best-of list | Moving, solar |
| trustanalytica.org | 8 | 6.7% | Best-of list | Moving, wedding, solar |
For a local business, this means citable writing on your own site is half the job. The other half is making sure the review sites and best-of lists for your industry carry accurate, specific facts about you, because those pages are what the AIs read most often before they recommend anyone.
The facts AI repeats are the specific, checkable ones
We searched the full text of all 120 answers for the kinds of facts this guide recommends adding. The counts come from a pattern search, so treat them as close estimates.
Which facts the AIs repeated about the businesses they recommended
Share of answers that mention each kind of fact at least once (pattern search of the full answer text)
60.8% of the 120 answers quoted a specific review count or star rating, roughly twice the share that mentioned years in business or awards.
- A review count or star rating60.8%
- Years in business or founding year31.7%
- Awards or "voted best"30%
- A price or price range27.5%
- Licensed, insured or BBB rating26.7%
- Certifications or accreditation25.8%
Source: Orkkid AI Recommendation Study, October 2026 (120 answers, 5 engines, 3 industries, 8 cities)
Specific numbers about reputation dominated. 60.8% of answers quoted a review count or a star rating, and the share varied sharply by engine: Claude did so in 95.8% of its answers, Perplexity in 91.7%, ChatGPT in 70.8%, Google AI Mode in 41.7% and Gemini in 4.2%. Years in business or a founding year appeared in 31.7% of answers and awards in 30%, with awards most common for wedding photographers (60%) and certifications most common for solar installers (50%).
Here is what that looks like in the answers themselves, quoted word for word:
- ChatGPT, Austin moving companies: "moveBuddha’s October 2026 Austin rankings list Black Tie Moving as its preferred Austin mover, with a 4.66/5 rating, 92% positive reviews, 12 years in business, and BBB A+ rating."
- ChatGPT, Charlotte solar installers: "Renu is based in Charlotte, has been around since 2010, and EnergySage lists it as an Elite Installer with a 4.7/5 rating from 294 reviews."
- Claude, Austin wedding photographers: "Matt Montalvo Photography is a multi-award winner with a 5.0 rating across 294 reviews on WeddingWire."
- ChatGPT, Denver wedding photographers, on Jesse La Plante: "Fearless Photographers lists him in Denver with 25 Fearless Awards, the highest Colorado total shown in its Colorado top-10 results."
- Gemini, Tampa solar installers, on Florida Power Services: "Operating in the Tampa Bay area since 2007, they are one of the longest-standing NABCEP-certified solar contractors in the region."
Notice that most of these facts are attributed to a third party: moveBuddha, EnergySage, WeddingWire, Fearless Photographers. The AIs repeat the founding year, review count, rating, award or certification that a review site already publishes. So put those exact facts on your own pages in quotable sentences ("Licensed and insured since 2008, rated 4.8 from 300 Google reviews"), and make sure the same numbers appear on your profiles on the review sites that matter in your industry.
How to test if your content will get cited by AI
Three quick tests you can run on any paragraph before publishing.
The lift test. Copy the paragraph, paste it into ChatGPT in a fresh chat, and ask: "If I asked you a question about [your topic], would you cite this paragraph as your source? Why or why not?" The model is candid about why it would or would not lift it.
The context-free test. Show the paragraph to someone who has not read the rest of the page. Ask them what it is about. If they need the surrounding context, the paragraph fails.
The first-sentence test. Read only the first sentence of every paragraph in your article. Does the result read like a reasonable summary? If yes, your paragraphs are anchored. If the first sentences are setups, the article is built for linear human reading, not for AI lift.
Is ChatGPT naming you, or your competitor?
Buyers now ask AI for a shortlist before they visit a single website. Find out whether your business is on it, and how to get there first.
Check my AI visibilityThe full workflow
For a new article or page rewrite, here is the order we follow at Orkkid.
- Decide the customer question the page should be cited for.
- Write a one-sentence answer to that question. This becomes the page's thesis.
- Outline 5 to 8 follow-up questions a customer would ask next. Each becomes an H2.
- Under each H2, write 1 to 3 paragraphs using the four properties above.
- Add specific numbers, dates, and named entities throughout.
- Add schema markup that matches the topic. Use our schema generator if needed.
- Publish. Add the URL to your llms.txt file.
- Wait 4 to 8 weeks. Check whether you are named with our AI citation checker.
This is the same writing-and-shipping cycle behind the results in our case studies. The technique is simple. The discipline of doing it every month is the rare part.
How we got this data
On 7 October 2026 we asked ChatGPT (gpt-5.5), Perplexity (sonar-pro), Gemini (gemini-3.8-flash) and Claude (claude-sonnet-5-5), all with web search on, plus Google AI Mode: "Who is the best {moving company | wedding photographer | solar installer} in {city}? Give me your top picks." We asked it in Denver, Austin, Phoenix, Charlotte, Seattle, Tampa, London and Toronto, which gave 120 answers naming 558 different businesses, and we matched each answer against the Google Maps top 20 for the same search (480 listings). Every cited domain was classified by type (review site, business website, best-of list and so on), with source shares counted once per domain per answer. The fact counts come from a pattern search of the full answer texts, so a few mentions may be missed or miscounted. Each prompt ran once per engine, so this is a snapshot: answers vary between runs, and the Maps top 20 reflects a city-centre search. The full method and results are in our AI recommendation study.
When a customer asks ChatGPT who to hire, whose name does AI give?
Your free audit tests the questions your customers ask across ChatGPT, Perplexity and Google AI, shows which businesses get named today, and lays out what it will take to add yours.
Tested across ChatGPT · Perplexity · Google AI Overviews · Gemini
Check my AI visibilityFrequently asked questions
Does paragraph length matter for AI citation?
Yes, indirectly. Paragraphs of 60 to 120 words are the most citable. Longer paragraphs (200+ words) tend to get partly lifted, which can produce odd quotes. Short paragraphs (under 30 words) often lack enough context to stand alone.
Should I use bullet lists or prose?
Both, depending on the question. Comparison and "what to look for" questions get cited from bullet lists. "What is" and "how does" questions get cited from prose. Most pages benefit from a mix.
How does this compare to traditional SEO writing?
Traditional SEO rewards articles. GEO rewards paragraphs. A 3,000-word article ranking #1 on Google might still produce zero AI citations if its paragraphs are not self-contained. And a single 80-word paragraph on a low-traffic page might get cited again and again if it answers a question well.
Do AI engines penalize SEO-optimized writing?
Not penalize, but ignore. Keyword-stuffed paragraphs do not get cited because they are not shaped like answers. Paragraphs that read naturally and contain specific information get cited whether or not they were "SEO-optimized".
If you want a citability review of your existing content, book a free AI visibility audit. We test the questions your buyers ask and show you which pages AI engines skip, and why.
For deeper context, read how ChatGPT recommends businesses and our Perplexity SEO guide.





